summaryrefslogtreecommitdiff
path: root/dix
diff options
context:
space:
mode:
authorPeter Hutterer <peter.hutterer@who-t.net>2009-02-12 15:05:57 +1000
committerPeter Hutterer <peter.hutterer@who-t.net>2009-02-13 14:41:57 +1000
commitcc696a2b7928ec497b6a0df3602fc70dec83a629 (patch)
treee6041564f818864794fb3b46c834a3133e6efc34 /dix
parent8364bf7374aa5f6d991700f0c02921dc6c638c9f (diff)
Doxygen-ify bits of events.c
Diffstat (limited to 'dix')
-rw-r--r--dix/events.c44
1 files changed, 24 insertions, 20 deletions
diff --git a/dix/events.c b/dix/events.c
index d34d7399a..14a0e5e57 100644
--- a/dix/events.c
+++ b/dix/events.c
@@ -107,7 +107,7 @@ of the copyright holder.
******************************************************************/
-/** @file
+/** @file events.c
* This file handles event delivery and a big part of the server-side protocol
* handling (the parts for input devices).
*/
@@ -163,9 +163,7 @@ typedef const char *string;
#include "enterleave.h"
-/**
- * Extension events type numbering starts at EXTENSION_EVENT_BASE.
- */
+/* Extension events type numbering starts at EXTENSION_EVENT_BASE. */
#define NoSuchEvent 0x80000000 /* so doesn't match NoEventMask */
#define StructureAndSubMask ( StructureNotifyMask | SubstructureNotifyMask )
#define AllButtonsMask ( \
@@ -246,30 +244,35 @@ InputInfo inputInfo;
/**
* syncEvents is the global structure for queued events.
+ *
* Devices can be frozen through GrabModeSync pointer grabs. If this is the
* case, events from these devices are added to "pending" instead of being
* processed normally. When the device is unfrozen, events in "pending" are
* replayed and processed as if they would come from the device directly.
- *
- * pending ... list of queued events
- * pendtail ... last event in list
- * replayDev ... The device to replay events for. Only set in AllowEvents, in
- * which case it is set to the device specified in the request.
- * replayWin ... the window the events are supposed to be replayed on. This
- * window may be set to the grab's window (but only when
- * Replay{Pointer|Keyboard} is given in the XAllowEvents
- * request.
- * playingEvents ... flag to indicate whether we're in the process of
- * replaying events. Only set in ComputeFreezes().
*/
static struct {
- QdEventPtr pending, *pendtail;
+ QdEventPtr pending, /**< list of queued events */
+ *pendtail; /**< last event in list */
+ /** The device to replay events for. Only set in AllowEvents(), in which
+ * case it is set to the device specified in the request. */
DeviceIntPtr replayDev; /* kludgy rock to put flag for */
+
+ /**
+ * The window the events are supposed to be replayed on.
+ * This window may be set to the grab's window (but only when
+ * Replay{Pointer|Keyboard} is given in the XAllowEvents()
+ * request. */
WindowPtr replayWin; /* ComputeFreezes */
+ /**
+ * Flag to indicate whether we're in the process of
+ * replaying events. Only set in ComputeFreezes(). */
Bool playingEvents;
TimeStamp time;
} syncEvents;
+/**
+ * The root window the given device is currently on.
+ */
#define RootWindow(dev) dev->spriteInfo->sprite->spriteTrace[0]
static xEvent* swapEvent = NULL;
@@ -277,6 +280,7 @@ static int swapEventLen = 0;
/**
* Convert the given event type from an XI event to a core event.
+ * @param[in] The XI 1.x event type.
* @return The matching core event type or 0 if there is none.
*/
int
@@ -298,8 +302,8 @@ XItoCoreType(int xitype)
}
/**
- * True if device owns a cursor, false if device shares a cursor sprite with
- * another device.
+ * @return true if the device owns a cursor, false if device shares a cursor
+ * sprite with another device.
*/
Bool
DevHasCursor(DeviceIntPtr pDev)
@@ -308,7 +312,7 @@ DevHasCursor(DeviceIntPtr pDev)
}
/*
- * Return true if a device is a pointer, check is the same as used by XI to
+ * @return true if a device is a pointer, check is the same as used by XI to
* fill the 'use' field.
*/
Bool
@@ -318,7 +322,7 @@ IsPointerDevice(DeviceIntPtr dev)
}
/*
- * Return true if a device is a keyboard, check is the same as used by XI to
+ * @return true if a device is a keyboard, check is the same as used by XI to
* fill the 'use' field.
*
* Some pointer devices have keys as well (e.g. multimedia keys). Try to not